Novel AI-informed spectroscopy and spectrometry methodologies for meat authenticity

: meat identification and meat evaluation

Abstract

With the increasing population and the demand for a healthy diet, ensuring the status of food has become an essential topic of discussion. Among all types of food, meat provides high-quality protein and a range of nutrients. While recognising its nutritional value, meat products are particularly plagued by fraud and safety concerns due to their diverse production processes and complex supply chains. Conventionally, analytical methods address perceived issues by focusing on one or a few components in an environmentally sensitive and time-consuming way. These concerns have become pressing when the demand for real-time monitoring and traceability of meat products increases. With these concerns in mind, spectroscopy and spectrometry methodologies are discovered as a new perspective to maintain real-time monitoring for regular operation and various places in the supply chain. Meanwhile, advanced chemometric techniques are essential to extract chemical information from generated data to address complex tasks regarding meat authentication. This thesis aims to apply and innovate chemometric methods with spectroscopy and mass spectrometry techniques to accurately identify and evaluate meat products. To be more specific, three techniques from the field of non-target analysis were utilised for meat authentication, focusing on fraud and quality issues: Near-Infrared (NIR), Hyperspectral Imaging (HSI), and Rapid Evaporation Ionisation Mass Spectrometry (REIMS). This thesis underscores the synergistic use of spectroscopy, spectrometry, and advanced chemometric methods to overcome meat fraud and safety issues. What was noticed is that three factors influence each experiment significantly: experimental design, sample preparation, and data analysis. Considering all these factors contributes to the verification of non-target techniques, which can transform the meat producer shifting from a reactive to a predictive approach in addressing meat issues.

Thesis embargoed until 31 July 2025.

Date of Award	Jul 2024
Original language	English
Awarding Institution	Queen's University Belfast
Sponsors	Queen's University & China Scholarship Council
Supervisor	Tassos Koidis (Supervisor), Saskia van Ruth (Supervisor) & Nigel Scollan (Supervisor)